It started innocently enough. Two acoustic consultants met at a conference cocktail reception to discuss acceptable noise levels in open-plan offices. The ambient level in the reception area was about 72 dBA — typical for a room full of engineers with complimentary wine.

'THE RECOMMENDED LEVEL IS NC-40,' said Dr. Patel, raising her voice to be heard. 'THAT'S ABOUT 45 dBA,' replied Professor Chen, louder still, because a nearby group had started laughing. The Lombard effect — the involuntary tendency to speak louder in the presence of background noise — had them firmly in its grip.

Within ten minutes, the two foremost experts on noise control were essentially shouting at each other about the importance of quiet environments. I know this because I was standing next to them with an SPL meter (occupational hazard) and their conversation was hitting 82 dBA at one meter.

'WE SHOULD REALLY FIND A QUIETER SPOT,' bellowed Dr. Patel, at a level that would trigger OSHA's action level. They moved to the corridor. Within moments, six other conversations followed them. The corridor was now louder than the reception had been.

The Lombard effect adds approximately 3-6 dB of vocal effort for every 10 dB increase in background noise. A room full of humans is a positive feedback loop with cocktails.

The Moral: The Lombard effect is real, measurable, and relentless. Use SonaVyx's SPL meter with time-history logging to observe it in real time — and remember that the single most effective noise control measure in any room full of people is giving them a reason to stop talking.